Research On The Evolution Of Knowledge Agent And Knowledge Service Networks In Agile Supply Chain

Supply chain is composed of many member enterprises with specific business knowledge, and knowledge insufficiency is inevitably existing in every company. For this reason, knowledge service networks with knowledge service as the core are established. On the platform of knowledge service, supply chain members conduct innovation of product and technology through obtaining knowledge resources needed. Thus to improve the knowledge service capability of knowledge absorption and transfer, and to improve the ability to optimize the configuration of knowledge resources. The differences of region, culture and management are overcame; the heterogeneous of information system, planning process and knowledge system are weakened. They seek more reasonable management under new organizational structure and competition mechanism, in order to increase the income, improve operation efficiency and realize the flexibility and agile of supply chain. Thus, this research is aiming at the following four aspects:(1)Research on the structure and knowledge service procedure of knowledge service network in agile supply chain. First, based on the theory of complex adaptive system, combined of the concept of knowledge service in agile supply chain and knowledge networks, knowledge service networks in agile supply chain is defined and its structure is analyzed as well as the features of knowledge agent. Then, knowledge service procedure and structure of knowledge service network in agile supply chain are described, the flow chart and structure chart are given. According to its knowledge service procedure and network structure, combining the basic theory of generalized stochastic Petri nets, GSPN of the knowledge service network in agile supply chain is defined and modular modeling with the modeling approach. The Markov chain isomorphic with the built model is constructed and steady-state probability distribution of the model is solved. At last, time performance and operation efficiency of the model are calculated and analyzed through model solution, so to improve the operation efficiency of knowledge service network in agile supply chain, and to realize the optimization of knowledge service procedure.(2)Research on the knowledge service capability of knowledge agent。Based on the dissipative structure theory, dissipative structure theory and entropy theory, first analyzes the fractal characteristics and evolutionary conditions of knowledge service network in agile supply chain which consists of fractal cells (knowledge agent). Knowledge service entropy is defined and the calculation formula is given. Based on that, the evolution direction distinguishing model of knowledge agent’s knowledge service capability of knowledge service network in agile supply chain is established with the distinguishing steps. Through example analysis and entropy calculation, the change of knowledge service capability could be judged, and the weak item of knowledge service capability could be identified, so that corresponding influence factors could be controlled and knowledge service capability of knowledge agent could be improved.Afterwards, evolutionary model is built according to the analysis and model of entropy change of knowledge service capability of knowledge agent. And then, evolutionary rate and entropy change process of four evolutionary stages are explored with the Logistic equation. At last, a Logistic growth curve is illustrated under two strategies which are delay strategy and sustainable evolution strategy. It is pointed out that the evolutionary process of knowledge service capability of knowledge agent could be controlled through controlling of threshold of state variables.(3)Research on the evolution of knowledge interaction of knowledge agent. First depict knowledge interaction procedure of knowledge agent of knowledge service network, and qualitatively analyze the feedback loop and influence factors combining with causal relationship graph. Furthermore, a system dynamics model is built and quantization relationships among variables are denoted by equations. After simulation and result analysis of critical variables is model validation and sensitivity test of equation parameters. The validation results show that the system dynamics model is valid for simulating the process of knowledge interaction. And sensitivity test inspect the influence of different intellectual capital investment on knowledge interaction and its effectiveness. The test results provide for knowledge agent of selecting intellectual capital investment program before knowledge interaction with theoretical basis and data analysis support.(4)Research on the evolution of knowledge service network in agile supply chain. Based on the complex adaptive system theory, the third-party knowledge service provider is introduced as knowledge service agent, together with other knowledge agents and environment agent, forming the structure of knowledge service network in agile supply chain. On the basis of putting forward the evolutionary procedure description of knowledge service in agile supply chain, the adaptive model of knowledge agent (KA), the selecting model of KA, the interacting model and the knowledge protecting model are setting as the sub-modules of KS-Net evolutionary model, and to analyze the evolution of knowledge service network in agile supply chain using the multi-objective decision model. It is pointed out that KS-Net in ASC, which is built up by the knowledge service agent through its knowledge service, is evolving under the multi-objective of maximizing the utility of all KAs and timeliness of knowledge service.On the other hand, integrated use of the grey theory and synergetic, this paper takes the co-evolution system dynamic model of knowledge service network in agile supply chain as object. On the basis of grey relational analysis of the main variables of sub-systems, a co-evolutionary grey system model of knowledge service network in agile supply chain is built with application of non-linear differential equation. And then the order parameters are identified through adiabatic elimination of the fast variables according to their relaxation coefficients. At last, the results show that co-evolution of sub-systems could be dominated by the control of order parameters and their determinants, thus to guide the entire system to a benign development.